In a conventional method for encoding a main picture comprised of video image data and a sub-picture comprised of only character information, such as captions superimposed on movie films, lyrics displayed on visual "karaoke" systems, telops displayed on TV screens, there is a method of encoding a sub-picture comprised of the character data after superimposing it onto a main picture comprised of video image data.
This encoding method does not require any special operation or circuit for displaying characters in a decoding operation and has no restriction on the shape or size of characters. It is possible to add decorations or other characters such as kana letters on or around the characters. However, in a low bit rate motion picture compression with a high data compression rate, deformations of the characters become conspicuous so that small characters become difficult to read. Furthermore, if one tries to reduce these deformations, the efficiency of the data compression deminishes.
Further, there is another encoding method which encodes a sub-picture using a code data format such as a format regulated by the Japanese Industrial Standard, separately from a main picture data. This encoding method has a feature that no deformation occurs in a decoded character, even when an image compression process has been applied to the character. However a character generator is necessary for displaying characters in its decoding operation, and the process of superimposing the sub-picture onto a main picture and the process of synchronizing them with each other have to be performed. In addition, the shape and size of the character are not flexible and it is not possible to display characters that cannot be generated by a character generator.
As described above, according to a conventional encoding system for a character image encoding/decoding apparatus, a method of encoding a sub-picture comprised of character data after superimposing it onto a main picture comprised of the video image data is used. However, the conventional method has a defect such that deformations of the characters become conspicuous so that small characters become difficult to read. Further, in a method for encoding character data using a code data format separately from a main picture, there are many restrictions on usable characters.